Epidemiologic risk factors for breast cancer implicate a role for steroid hormones in breast cancer etiology. While there is strong evidence of a positive association between pre-diagnostic levels of serum estrogens and androgens and risk of breast cancer in postmenopausal women, there is little data for premenopausal women. Most published studies are limited by small numbers of premenopausal breast cancer cases and lack of control of variation in serum sex steroid hormone levels related to the ovarian cycle. In addition to steroid hormones, new evidence implicates insulin-like-growth factor (IGF)-I and insulin in the etiology of premenopausal breast cancer. Besides the direct effect on proliferation and apoptosis inhibition exert by insulin and IGF-I on tumor cells, the etiological rote of those variables may also be explained by their effect on tumor promotion by up regulation of ovarian steroid secretion. We propose here to evaluate the association of breast cancer with the endogenous steroid hormones (estradiol, testosterone and progesterone) and with IGF-I and insulin in premenopausal women using data from the Italian ORDET prospective cohort study. In that study, particular care was taken during participant recruitment to control several potential sources of hormone biological variability both by inclusion criteria and by highly standardized conditions at blood sample collection. In addition, we propose to evaluate whether the association of androgens and estrogens is explained, at least in part, by an underlying hyperinsulinemia-increased IGF-1 bioavailability status. The study will be conducted as a nested case-control study. 5,794 healthy volunteer premenopausal women, residents in Varese province, an area covered by the Lombardy Cancer Registry, were enrolled between 1987 and 1992. We expect 302 breast cancer cases among premenopausal cohort members during the 17 year-follow-up period. Four controls will be matched to each case on age, menopausal status at diagnosis, recruitment center and recruitment period for a total number of 1,208 control subjects. This study will provide important data on risk factors that are likely key to the development of premenopausal breast cancer.